Recently we showed an integral
epidermal growth factor receptor (EGFR)-E2F3a signaling path, in which E2F3a was found to be essential in EGFR-mediated proliferation in
ovarian cancer cells. The present work evaluates the clinical relevance of this novel axis and of E2F3a itself in a large set of 130
ovarian cancer specimens. For this purpose E2F3a and its counterpart, E2F3b, were measured by RT-PCR and activated EGFR was assessed by immunohistochemistry. When compared with healthy control tissue, both E2F3
isoforms were overexpressed in the
cancers, but only E2F3a expression correlated with
tumor stage (ρ=0.349, P=0.0001) and residual disease (ρ=0.254, P=0.004). Univariate survival analyses showed E2F3a and activated EGFR to be associated with poor PFS and OS. Furthermore, a strong, positive correlation between activated EGFR and E2F3a expression was shown (P=0.0001). We further identified two EGFR-independent mechanisms that regulate E2F3a expression, namely one, acting by promoter methylation of miR-34a, which by its physical interaction with E2F3a transcripts causes their degradation, and the second based on 6p22 gene locus amplification. MiRIDIAN-based knockdown and induction of miR-34a evidenced a direct regulatory link between miR-34a and E2F3a, and the
tumor-suppressive character of miR-34a was documented by its association with improved survival. Although, 6p22 gene locus amplification was detected in a significant number of
ovarian cancer specimens, 6p22 ploidy was not relevant in predicting survival. In Cox regression analysis, E2F3a, but not activated EGFR or miR-34a expression, retained independent prognostic significance (PFS: hazards ratio 3.785 (1.326-9.840), P=0.013; OS: hazards ratio 4.651 (1.189-15.572), P=0.013). These clinical findings highlight the relevance of E2F3a in the biology of
ovarian cancer. Moreover, identification of EGFR-independent mechanisms in E2F3a control can be helpful in explaining the non-responsiveness of therapeutic EGFR targeting in
ovarian cancer.